File size: 5,401 Bytes
4c8c217
52b8278
74d32cc
0ff2378
 
7d9fec0
0ff2378
 
 
7d9fec0
 
 
98f8b95
49b8bd6
b594097
7d9fec0
1dd3a02
7d9fec0
1dd3a02
b594097
986e2b7
 
7d9fec0
 
 
 
 
 
 
 
 
986e2b7
0ff2378
 
 
 
 
 
 
7d9fec0
 
0ff2378
 
 
 
c790556
7d9fec0
c790556
0ff2378
 
 
 
 
 
 
 
 
 
 
 
1dd3a02
c790556
0ff2378
c790556
0ff2378
 
 
 
 
 
 
 
 
 
 
8a6e9d6
a9a1953
 
 
 
 
 
 
 
 
 
49b8bd6
 
 
 
 
 
4f92556
49b8bd6
 
 
 
2f9704c
c790556
 
0ff2378
 
 
 
 
 
 
 
 
885ec28
b1d9a7f
 
885ec28
 
0ff2378
4c8c217
1dd3a02
 
b0d0303
4c8c217
 
81a4020
6b80b65
81a4020
 
1572555
52b8278
 
0ff2378
 
 
 
 
 
 
c306f1c
986e2b7
 
52b8278
 
 
81a4020
8c67f60
 
52b8278
 
 
 
 
 
 
 
 
 
6ad6f7e
 
 
 
6b80b65
4c8c217
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
import streamlit as st
from langchain_core.messages import AIMessage, HumanMessage
from langchain_community.document_loaders import WebBaseLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_community.vectorstores import Chroma
# from langchain_openai import OpenAIEmbeddings, ChatOpenAI
from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
from langchain.chains import create_history_aware_retriever, create_retrieval_chain
from langchain.chains.combine_documents import create_stuff_documents_chain
from langchain_community.embeddings import HuggingFaceBgeEmbeddings
from langchain_community.llms import CTransformers
from ctransformers import AutoModelForCausalLM
from langchain.llms import HuggingFaceHub
from transformers import AutoModelForCausalLM, AutoTokenizer
import os
# from dotenv import load_dotenv

# load_dotenv()



def get_vector_store_from_url(url):
    model_name = "BAAI/bge-large-en"
    model_kwargs = {'device': 'cpu'}
    encode_kwargs = {'normalize_embeddings': False}
    embeddings = HuggingFaceBgeEmbeddings(
        model_name=model_name,
        model_kwargs=model_kwargs,
        encode_kwargs=encode_kwargs
    )
    
    loader = WebBaseLoader(url)
    document = loader.load()
    
    # split the document into chunks
    text_splitter = RecursiveCharacterTextSplitter()
    document_chunks = text_splitter.split_documents(document)
    
    # create a vectorstore from the chunks
    # vector_store = Chroma.from_documents(document_chunks, OpenAIEmbeddings())
    vector_store = Chroma.from_documents(document_chunks, embeddings)

    return vector_store
    

def get_context_retriever_chain(vector_store,llm):
    # llm = ChatOpenAI()
    llm = llm
    retriever = vector_store.as_retriever()
    
    prompt = ChatPromptTemplate.from_messages([
      MessagesPlaceholder(variable_name="chat_history"),
      ("user", "{input}"),
      ("user", "Given the above conversation, generate a search query to look up in order to get information relevant to the conversation")
    ])
    
    retriever_chain = create_history_aware_retriever(llm, retriever, prompt)
    
    return retriever_chain
    

def get_conversational_rag_chain(retriever_chain,llm): 
    
    llm=llm
    
    prompt = ChatPromptTemplate.from_messages([
      ("system", "Answer the user's questions based on the below context:\n\n{context}"),
      MessagesPlaceholder(variable_name="chat_history"),
      ("user", "{input}"),
    ])
    
    stuff_documents_chain = create_stuff_documents_chain(llm,prompt)
    
    return create_retrieval_chain(retriever_chain, stuff_documents_chain)

def get_response(user_input):
    # llm = CTransformers(
    #                 # model = "TheBloke/Mistral-7B-Instruct-v0.2-GGUF",
    #                 model= "TheBloke/Llama-2-7B-Chat-GGUF",
    #                 model_file = "llama-2-7b-chat.Q3_K_S.gguf",
    #                 model_type="llama",
    #                 max_new_tokens = 300,
    #                 temperature = 0.3,
    #                 lib="avx2", # for CPU
    # )

    model_name = "TinyLlama/TinyLlama-1.1B-Chat-v1.0"
    # llm = HuggingFaceHub(
    #     repo_id=llm_model, 
    #     model_kwargs={"temperature": 0.3, "max_new_tokens": 250, "top_k": 3}
    # )

    llm = transformers.AutoModelForCausalLM.from_pretrained(
                model_name,
                trust_remote_code=True,
                torch_dtype=torch.bfloat16,
                device_map='auto'
    )
    retriever_chain = get_context_retriever_chain(st.session_state.vector_store,llm)
    conversation_rag_chain = get_conversational_rag_chain(retriever_chain,llm)
    
    response = conversation_rag_chain.invoke({
        "chat_history": st.session_state.chat_history,
        "input": user_query
    })
    
    return response['answer']

    
# app config
st.set_page_config(page_title= "Chat with Websites", page_icon="🤖")
st.title("Chat with Websites")





#sidebar
with st.sidebar:
    st.header("Settings")
    website_url = st.text_input("Website URL")
    # openai_apikey = st.text_input("Enter your OpenAI API key")

if (website_url is None or website_url == ""):
    st.info("Please ensure if website URL is entered")
    

else:
    
    if "chat_history" not in st.session_state:
        st.session_state.chat_history = [ 
        AIMessage(content = "Hello, I am a bot. How can I help you"),
        ]

    if "vector_store" not in st.session_state:
        st.session_state.vector_store = get_vector_store_from_url(website_url)

    
    #user_input
    user_query = st.chat_input("Type your message here...")
    if user_query is not None and user_query !="":
        response = get_response(user_query)
        st.session_state.chat_history.append(HumanMessage(content=user_query))
        st.session_state.chat_history.append(AIMessage(content=response))
        
                
    #conversation
    for message in st.session_state.chat_history:
        if isinstance(message, AIMessage): # checking if the messsage is the instance of an AI message
            with st.chat_message("AI"):
                st.write(message.content)
        elif isinstance(message, HumanMessage): # checking if the messsage is the instance of a Human
            with st.chat_message("Human"):
                st.write(message.content)